Bacillus subtilis, a medically and industrially important soil growing bacterium, is well known for its rich stress response spectrum. Among the many ingenious programs employed by B. subtilis to cope with stress are its ability to take up extracellular DNA, differentiateinto spores, synthesize degradative enzymes, and become motile. It has long been observed that B. subtilis exhibits epigenetic stress response diversification: that is, in a genetically isoclonal population of cells under identical environmental conditions, some cells will embark on the 'selected' stress response, while the remainder activate alternative pathways. Laboratory conditions are designed to militate against this clonal diversity, an undesirable complication in many experiments and industrial applications. However, either phenotypic noise is spurious, existing only because there is no strong selection for a uniform response; or diversification is controlled by the cell and serves an evolutionary purpose. This proposal is based on this latter view. We base this hypothesis on a growing body of evidence that stress response diversification in microbes can be an adaptive response to an unpredictable environment, and by our preliminary modeling studies. Here we propose a directed program to quantify the probabilistic decision-making across multiple stress response pathways in B. subtilis, and to elucidate the network features giving rise to switching and diversification. More specifically, we propose to systematically measure the fraction of B. subtilis cells committing to all possible combinations of spore formation, competence for DNA transformation, subtilisin synthesis, and motility, in response to a series of stress conditions designed to exercise each part of the integrated network controlling these processes. Because our experimental design calls for a series of stressors, rather than single stressors, these data will also begin to quantify environmental memory in the circuitry. We plan to collect data at an average population level, a subpopulation level, and a single cell level, all pertaining to the expression patterns of the four stress responses. This data set,made unique by the coordinated measurements across multiple modalities, will be analyzed through formal comparison to a mathematical model. This study is the first effort to systematically measure, model, and explain coordinated diversification across multiple stress responses, and will form the foundation for future studies on the physiological significance of diversification in microbes.